Aspects of cell calculations in deterministic reactor core analysis

Τhe capability of achieving optimum utilization of the deterministic neutronic codes is very important, since, although elaborate tools, they are still widely used for nuclear reactor core analyses, due to specific advantages that they present compared to Monte Carlo codes. The user of a deterministic neutronic code system has to make some significant physical assumptions if correct results are to be obtained. A decisive first step at which such assumptions are required is the one-dimensional cell calculations, which provide the neutronic properties of the homogenized core cells and collapse the cross sections into user-defined energy groups. One of the most crucial determinations required at the above stage and significantly influencing the subsequent three-dimensional calculations of reactivity, concerns the transverse leakages, associated to each one-dimensional, user-defined core cell. For the appropriate definition of the transverse leakages several parameters concerning the core configuration must be taken into account. Moreover, the suitability of the assumptions made for the transverse cell leakages, depends on earlier user decisions, such as those made for the core partition into homogeneous cells. In the present work, the sensitivity of the calculated core reactivity to the determined leakages of the individual cells constituting the core, is studied. Moreover, appropriate assumptions concerning the transverse leakages in the one-dimensional cell calculations are searched out. The study is performed examining also the influence of the core size and the reflector existence, while the effect of the decisions made for the core partition into homogenous cells is investigated. In addition, the effect of broadened moderator channels formed within the core (e.g. by removing fuel plates to create space for control rod hosting) is also examined. Since the study required a large number of conceptual core configurations, experimental data could not be available for comparison. For this purpose, corresponding reactivity results obtained by detailed geometry/high statistics Monte Carlo calculations, performed by the well documented neutronic code TRIPOLI, were utilized. It was concluded that the assumptions made for the transverse leakages in the one-dimensional cell calculations affect significantly the core reactivity computation. Small modifications in the parameters which determine the transverse cell leakages can induce keff modifications of the order of 103 pcm. Significant factors for the suitability of the adopted assumptions were also shown to be the core size and the inclusion of several components, such as reflector or internal irradiation channels.